Educational attainment and response to HAART during initial therapy for HIV-1 infection B Linda G. Marc a,b,c, * , Marcia A. Testa c , Alexander M. Walker c , Gregroy K. Robbins d , Robert W. Shafer e , Norman B. Anderson c,f , Lisa F. Berkman c , for the ACTG Data Analysis Concept Sheet Study Team a Cornell HIV Clinical Trials Unit, New York, NY, USA b Department of Psychiatry, Weill Medical College of Cornell, White Plains, NY, USA c Harvard School of Public Health, Boston, MA, USA d Massachusetts General Hospital, Boston, MA, USA e Stanford University Medical Center, Stanford, CA, USA f American Psychological Association, Washington, DC, USA Received 24 September 2006 Abstract Objective: Previous research has demonstrated an association between educational attainment (EA) and negative physical and psychological outcomes. This study investigated whether EA is associated with regimen failure during initial therapy with highly active antiretroviral treatment (HAART) and whether adherence self-efficacy (ASE), a coping resource, moderates the relationship between EA and regimen failure. Methods: A secondary analysis of AIDS Clinical Trial Group Protocol 384, an international, multi- center, randomized, partially double-blinded trial, included 799 male and 181 female antiretroviral-naRve subjects (age, 37.0F9.5 years). Participants were recruited from 1998 to 1999 and followed for a median of 2.3 years across 81 centers. The dependent variable was btime to first regimen failure.Q Covariates include baseline HIV-1 log 10 RNA and CD4 + counts, self-reported adherence, study site, ASE, age, sex, race, treatment assignment, and baseline use of nonantiretroviral medications. Results: ASE significantly moder- ated the relationship between EA and regimen failure. Results showed that for every 10-unit increase in ASE, individuals with bless than high schoolQ education had a 17% reduction in regimen failure (hazard ratio=0.83; 95% confidence interval=0.70–0.98) when compared to the reference group bcollege/graduate,Q even after adjusting for baseline factors known to contribute to regimen failure. The time to first regimen failure was shorter with decreasing EA, trending toward significance ( P=.08). Conclusions: There is a social gradient in HAART effectiveness, and ASE reduces the deleterious effects of lower EA on regimen failure. We recommend designing controlled interventions to evaluate the effectiveness of programs that increase ASE prior to initiation with HAART, particularly for those with lower EA. D 2007 Elsevier Inc. All rights reserved. Keywords: Adherence self-efficacy; Antiretroviral; Educational attainment; HIV/AIDS; Socioeconomic status; Regimen failure Introduction The success or failure of antiretroviral therapy is not a purely biologic phenomenon [1–9] but rather the product of a complex interaction between biologic, behavioral, and societal factors [10–13]. While there is scarce evidence to support a social gradient in antiretroviral effectiveness, education has been known to be significantly related to both CD4 rate and change in HIV viral load [14]. There also have 0022-3999/07/$ – see front matter D 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.jpsychores.2007.04.009 B During the study period, Dr. Marc received consultancy fees from Aventis, Behavioral Science International LLC, GlaxoSmithKline, Parke- Davis, and Phase V Technologies. 4 Corresponding author. Department of Psychiatry, Weill Medical College of Cornell, 21 Bloomingdale Road, Unit 2 South, White Plains, NY 10605, USA. Tel.: +1 646 541 6650. E-mail address: linda.marc@post.harvard.edu (L.G. Marc). Journal of Psychosomatic Research 63 (2007) 207 – 216